Multidimensional Acoustic–Prosodic Quantification Framework Using Unscripted Speech for Autism Spectrum Disorder Identification
Autism Research · 2026-04 · 19 (4)
DOI: 10.1002/aur.70206 ↗Kurz gefasst
Fachartikel in der Zeitschrift Autism Research, online erschienen am 25.02.2026. Verfasst von Minghao Du und 9 weiteren Autor:innen. Beteiligte Einrichtung: Medical School, Tianjin University, Tianjin, China. Hauptthemen laut MEDLINE: Autism Spectrum Disorder, Speech Acoustics, Speech, Speech Production Measurement. Die Arbeit nennt 55 Literaturangaben. Gefördert von: the Natural Science Foundation of Tianjin, Autonomous Project of Haihe Laboratory of Brain-Computer Interaction and Human-Machine Integration, Natural Science Foundation of Tianjin Municipality.
Abstract (Originalsprache)
Although clinical observations have noted early speech abnormalities in children with autism spectrum disorder (ASD), automatic speech‐based detection remains challenging. This is primarily due to the reliance on scripted tasks, which younger children often struggle to complete and which are not generalizable to large‐scale, non‐clinical screening. To address this, we developed an unscripted speech‐based framework to quantify atypical acoustic–prosodic patterns for automatic ASD identification in naturalistic interactions. It processes free‐flowing conversations, extracts multidimensional acoustic features from the time and frequency domains, and models ASD‐related prosodic patterns for classification. For evaluation, we collected spontaneous speech from 88 children with ASD (3–10 years) and 82 typically developing (TD) children (3–9 years) during naturalistic interactions on daily topics (e.g., toys, animated movies, storybook reading). Group comparisons revealed atypical prosodic patterns in ASD, including reduced speech continuity, speech rate, and Formant 3, alongside increased zero‐crossing rate, pitch, pitch variability, and Formant 1 (all p < 0.01). Using these features, a linear discriminant analysis classifier achieved robust performance (accuracy = 0.85 ± 0.07, F1 = 0.86 ± 0.07). Further analyses indicated no significant gender interaction ( p > 0.05), but a pronounced effect of speech context ( p < 0.01), with atypical patterns being more evident in open‐ended dialogues than in text‐guided settings. Moreover, these patterns correlated with clinical scores ( p < 0.05), particularly language ability, demonstrating the framework's utility for assessing ASD severity. These findings underscore the importance of analyzing unscripted speech to capture atypical prosodic patterns and provide a basis for large‐scale ASD screening outside clinical settings.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Medical School, Tianjin University, Tianjin, China.
- State Key Laboratory of Advanced Medical Materials and Medical Devices, Tianjin University, Tianjin, China.
- Children's Hospital, Tianjin University, Tianjin, China.
- School of Biomedical Engineering and Technology, Tianjin Medical University, Tianjin, China.
- Haihe Laboratory of Brain-Computer Interaction and Human-Machine Integration, Tianjin, China.
ORCID-Profile: Shuang Liu
Förderung
- the Natural Science Foundation of Tianjin
- Autonomous Project of Haihe Laboratory of Brain-Computer Interaction and Human-Machine Integration
- Natural Science Foundation of Tianjin Municipality
Zitieren
APA
Du, M., Shi, P., Liu, Z., Lu, X., Cao, L., Liu, B., Liu, X., Liu, W., Liu, S., & Ming, D. (2026). Multidimensional Acoustic–Prosodic Quantification Framework Using Unscripted Speech for Autism Spectrum Disorder Identification. Autism Research, 19(4). https://doi.org/10.1002/aur.70206
Vancouver
Du M, Shi P, Liu Z, Lu X, Cao L, Liu B, et al. Multidimensional Acoustic–Prosodic Quantification Framework Using Unscripted Speech for Autism Spectrum Disorder Identification. Autism Research. 2026;19(4). doi:10.1002/aur.70206
BibTeX
@article{Du2026,
author = {Du, Minghao and Shi, Ping and Liu, Zehao and Lu, Xiaoyao and Cao, Luling and Liu, Beibei and Liu, Xiaoya and Liu, Wei and Liu, Shuang and Ming, Dong},
title = {Multidimensional Acoustic–Prosodic Quantification Framework Using Unscripted Speech for Autism Spectrum Disorder Identification},
journal = {Autism Research},
year = {2026},
volume = {19},
number = {4},
doi = {10.1002/aur.70206}
}
EUVIMED